WASHINGTON — The biggest banks are considering whether to abandon an effort to use their own internal models to help set capital requirements as they examine ways to streamline the post-crisis compliance landscape.
The largest banks fought hard to create the so-called advanced approaches framework in 2007 as part of Basel II, which let institutions with more than $250 billion of assets or $10 billion of on-balance-sheet international exposure develop internal models for assessing risk-based capital compliance.
The program was intended to let banks have a better chance to demonstrate their compliance — and potentially lower capital requirements — but several banking industry insiders are wondering if it's worth the effort, given other safeguards in place that prevent them from holding lower capital.
"That's one of the debates that's going on right now at the biggest banks," said Wayne Abernathy, executive vice president of financial institutions policy and regulatory affairs at the American Bankers Association. "I would say it has not yet reached a conclusion, but they're debating amongst themselves the value of the advanced approaches. Some say the CCAR [the Federal Reserve's stress test exercise] really does anything that they want the advanced approaches to do, and it's made them irrelevant."
Karen Shaw Petrou, managing partner at Federal Financial Analytics, said the advanced approaches program is made less useful if risk-weighted assets are not the binding capital constraint that banks face. A recently enacted supplementary leverage ratio limits the largest banks' ability to hold less capital. As a result, going through the costly process of developing granular models to assess risk-based capital is a pointless exercise.
"These advanced approaches do better track risk, but they're very expensive to do, and why would a bank invest in all of that when the binding constraint is the leverage capital ratio anyhow?" Petrou said.
Under the Basel framework, certain banks must begin the process of establishing their advanced approaches models if they exceed the asset thresholds, but smaller institutions may also opt in to the program if they choose. The Fed and the bank then must undergo so-called parallel runs of the bank's model and the Fed's standard model simultaneously over four consecutive quarters.
To date, 10 bank holding companies have completed their parallel runs: Bank of America, Wells Fargo, Citigroup, Goldman Sachs, JPMorgan Chase, Bank of New York Mellon, Morgan Stanley, Northern Trust, U.S. Bancorp and State Street. Many of those firms' individual depository subsidiaries also are enrolled and have conducted their own parallel runs with their primary regulators. Another three bank holding companies — American Express, PNC and Capital One — are in the process of conducting parallel runs.
Petrou said the banks pushed for the provision in Basel II in part because the risk weighting in Basel I accords was so primitive, effectively treating secured loans from reputable borrowers the same as far riskier loans. That structure incentivized banks to engage in more risk taking, since the capital required to backstop a safe loan is the same as that required to backstop a risky one while the returns are far different.
"They wanted it in part because Basel I was stupid," Petrou said. "That created a significant amount of structural risk-taking.
But it isn't just the Fed's Comprehensive Capital Analysis and Review annual test that's making the advanced approach potentially irrelevant. As part of the Dodd-Frank Act, Congress enacted in 2010 the Collins amendment, which requires that in instances where there are two separate ways of calculating capital compliance, the more stringent standard should apply. The amendment is meant to reduce regulatory arbitrage, but it also effectively renders advanced approaches models moot, since the Fed's more conservative models will always apply.
The Federal Reserve declined to comment specifically on the future of the advanced approaches framework. But Gov. Daniel Tarullo has said in the past that, while internal modeling is an important element of safety and soundness in the contexts of the stress tests, "we should consider discarding the [internal ratings-based] approach to risk-weighted capital requirements" because between the Collins amendment and the leverage ratio, ratings-based models have "little useful role to play."
Abernathy said giving the advanced approaches regime another look could be beneficial to the Fed, since the agency spends manpower examining the banks' advanced approaches models on an ongoing basis.
"The debate that's taking place is [whether] they really provide value for anybody, and whether they are worth the incredible superstructure that's being built on top of them," Abernathy said. "The [banks] that are currently still the advocates of advanced approaches are the ones who have invested the most in them, and they've got something of a stake in [them]. But on the other hand, there are others asking the question, 'It cost us millions to create them, it still costs us millions to run them, and what are we getting out of them?' "